US10140646B2 - System and method for analyzing features in product reviews and displaying the results - Google Patents
System and method for analyzing features in product reviews and displaying the results Download PDFInfo
- Publication number
- US10140646B2 US10140646B2 US14/846,501 US201514846501A US10140646B2 US 10140646 B2 US10140646 B2 US 10140646B2 US 201514846501 A US201514846501 A US 201514846501A US 10140646 B2 US10140646 B2 US 10140646B2
- Authority
- US
- United States
- Prior art keywords
- features
- product
- feature
- reviews
- score
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active, expires
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0631—Item recommendations
-
- G06F17/211—
-
- G06F17/214—
-
- G06F17/2235—
-
- G06F17/2765—
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/10—Text processing
- G06F40/103—Formatting, i.e. changing of presentation of documents
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/10—Text processing
- G06F40/103—Formatting, i.e. changing of presentation of documents
- G06F40/109—Font handling; Temporal or kinetic typography
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/10—Text processing
- G06F40/12—Use of codes for handling textual entities
- G06F40/134—Hyperlinking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F7/00—Methods or arrangements for processing data by operating upon the order or content of the data handled
- G06F7/06—Arrangements for sorting, selecting, merging, or comparing data on individual record carriers
- G06F7/08—Sorting, i.e. grouping record carriers in numerical or other ordered sequence according to the classification of at least some of the information they carry
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0282—Rating or review of business operators or products
Definitions
- This disclosure relates generally to reviews, and relates more particularly to a system and method for more thoroughly analyzing reviews.
- FIG. 1 illustrates a front elevation view of a computer system that is suitable for implementing an embodiment of the system disclosed in FIG. 3 ;
- FIG. 2 illustrates a representative block diagram of an example of the elements included in the circuit boards inside a chassis of the computer system of FIG. 1 ;
- FIG. 3 illustrates a flowchart for an exemplary method of calculating a feature score, according to an embodiment
- FIG. 4 illustrates a table of reviews and features, according to an embodiment
- FIG. 5 illustrates a table of reviews and features, according to an embodiment
- FIG. 6 illustrates a table of reviews and features, according to an embodiment
- FIG. 7 is an exemplary screen shot illustrating the presentation of feature scores
- FIG. 8 is a block diagram of a system arranged to perform various tasks according to an embodiment
- FIG. 9 is a flowchart illustrating a method for annotating reviews
- FIG. 10 illustrates an exemplary screen shot according to an embodiment
- FIG. 11 is an exemplary screen shot according to an embodiment
- FIG. 12 is an exemplary screen shot according to an embodiment
- FIG. 13 is a flowchart illustrating a method for annotating reviews
- FIG. 14 is an exemplary screen shot according to an embodiment
- FIG. 15 is a flowchart illustrating a method for creating a tag cloud.
- FIG. 16 is a block diagram of a system arranged to perform various tasks according to an embodiment.
- Couple should be broadly understood and refer to connecting two or more elements mechanically and/or otherwise. Two or more electrical elements might be electrically coupled together, but not be mechanically or otherwise coupled together. Coupling might be for any length of time, e.g., permanent or semi-permanent or only for an instant. “Electrical coupling” and the like should be broadly understood and include electrical coupling of all types. The absence of the word “removably,” “removable,” and the like near the word “coupled,” and the like does not mean that the coupling, etc. in question is or is not removable.
- two or more elements are “integral” if they are comprised of the same piece of material. As defined herein, two or more elements are “non-integral” if each is comprised of a different piece of material.
- “approximately” can, in some embodiments, mean within plus or minus ten percent of the stated value. In other embodiments, “approximately” can mean within plus or minus five percent of the stated value. In further embodiments, “approximately” can mean within plus or minus three percent of the stated value. In yet other embodiments, “approximately” can mean within plus or minus one percent of the stated value.
- a method can comprise: producing a feature set by creating a list of features for a product; analyzing user-generated content to find one or more mentions of each feature of the feature set; analyzing each of the one or more mentions of each feature of the feature set to determine a total polarity of each of the one or more mentions; calculating a feature score using the total polarity of each feature of the feature set; and displaying the feature score of each feature to a user.
- a system can comprise: one or more processing modules; and one or more non-transitory storage modules storing computing instructions configured to run on the one or more processing modules and perform the acts of producing a feature set by creating a list of a features for a product; analyzing user-generated content to find one or more mentions of each feature of the feature set; analyzing each of the one or more mentions of each feature of the feature set to determine a total polarity of each of the one or more mentions; calculating a feature score using the total polarity of each feature of the feature set; and displaying the feature score of each feature to a user.
- a method can comprise: analyzing one or more reviews in a set of reviews to determine a set of features; matching the set of features with the set of reviews to create an annotated set of reviews, wherein each review is matched with one or more annotations; for each annotation in each review in the annotated set of reviews, marking the annotation as positive for a feature if a feature score is greater than a first predetermined score; for each annotation in each review in the annotated set of reviews, marking the annotation as negative for a feature if a feature score is less than a second predetermined score; and transmitting for display one or more reviews of the annotated set of reviews; wherein: each of the features of the set of features represents a different aspect of a product being reviewed; and the feature score of a feature is calculated by determining a total polarity of reviews.
- a system can comprise: one or more processing modules; and one or more non-transitory storage modules storing computing instructions configured to run on the one or more processing modules and perform the acts of analyzing one or more reviews in a set of reviews to determine a set of features; matching the set of features with the set of reviews to create an annotated set of reviews, wherein each review is matched with one or more annotations; for each annotation in each review in the annotated set of reviews, marking the annotation as positive for a feature if a feature score is greater than a first predetermined score; for each annotation in each review in the annotated set of reviews, marking the annotation as negative for a feature if a feature score is less than a second predetermined score; and transmitting for display one or more reviews of the annotated set of reviews; wherein: each of the features of the set of features represents a different aspect of a product being reviewed; and the feature score of a feature is calculated by determining a total polarity of reviews.
- a method can comprise: analyzing one or more reviews in a set of reviews to determine a set of features analyzed in the set of reviews; for each review in the set of reviews, noting each feature of the set of features analyzed in the review; receiving a request from a user indicating a desire to view a selected feature in the set of features; determining which reviews in the set of reviews are noted with the selected feature and placing those reviews in a subset of reviews; and transmitting for display a summary of the subset of reviews.
- a system can comprise: a user input device; a display device; one or more processing modules; and one or more non-transitory storage modules storing computing instructions configured to run on the one or more processing modules and perform the acts of: analyzing one or more reviews in a set of reviews to determine a set of features analyzed in the set of reviews; for each review in the set of reviews, noting each feature of the set of features analyzed in the review; receiving a request from a user indicating a desire to view a selected feature in the set of features; determining which reviews in the set of reviews are noted with the selected feature and placing those reviews in a subset of reviews; and transmitting for display a summary of the subset of reviews
- FIG. 1 illustrates an exemplary embodiment of a computer system 100 , all of which or a portion of which can be suitable for implementing the techniques described herein.
- a chassis 102 and its internal components
- one or more elements of computer system 100 e.g., a refreshing monitor 106 , a keyboard 104 , and/or a mouse 110 , etc.
- a refreshing monitor 106 can also be appropriate for implementing the techniques described herein.
- Computer system 100 comprises chassis 102 containing one or more circuit boards (not shown), a Universal Serial Bus (USB) port 112 , a Compact Disc Read-Only Memory (CD-ROM), Digital Video Disc (DVD) drive, or Blu-Ray drive 116 , and a hard drive 114 .
- a representative block diagram of the elements included on the circuit boards inside chassis 102 is shown in FIG. 2 .
- a central processing unit (CPU) 210 in FIG. 2 is coupled to a system bus 214 in FIG. 2 .
- the architecture of CPU 210 can be compliant with any of a variety of commercially distributed architecture families.
- system bus 214 also is coupled to a memory storage unit 208 , where memory storage unit 208 comprises both read only memory (ROM) and random access memory (RAM).
- ROM read only memory
- RAM random access memory
- Non-volatile portions of memory storage unit 208 or the ROM can be encoded with a boot code sequence suitable for restoring computer system 100 ( FIG. 1 ) to a functional state after a system reset.
- memory storage unit 208 can comprise microcode such as a Basic Input-Output System (BIOS) or Unified Extensible Firmware Interface (UEFI).
- BIOS Basic Input-Output System
- UEFI Unified Extensible Firmware Interface
- the one or more memory storage units of the various embodiments disclosed herein can comprise memory storage unit 208 , a USB-equipped electronic device, such as, an external memory storage unit (not shown) coupled to universal serial bus (USB) port 112 ( FIGS. 1-2 ), hard drive 114 ( FIGS. 1-2 ), and/or CD-ROM, DVD drive, or Blu-Ray drive 116 ( FIGS. 1-2 ).
- the one or more memory storage units of the various embodiments disclosed herein can comprise an operating system, which can be a software program that manages the hardware and software resources of a computer and/or a computer network.
- the operating system can perform basic tasks such as, for example, controlling and allocating memory, prioritizing the processing of instructions, controlling input and output devices, facilitating networking, and managing files.
- Some examples of common operating systems can comprise various versions/distributions of Microsoft® Windows® operating system (OS), Apple® OS X, UNIX® OS, and Linux® OS.
- processor and/or “processing module” means any type of computational circuit, such as but not limited to a microprocessor, a microcontroller, a controller, a complex instruction set computing (CISC) microprocessor, a reduced instruction set computing (RISC) microprocessor, a very long instruction word (VLIW) microprocessor, a graphics processor, a digital signal processor, or any other type of processor or processing circuit capable of performing the desired functions.
- CISC complex instruction set computing
- RISC reduced instruction set computing
- VLIW very long instruction word
- the one or more processors of the various embodiments disclosed herein can comprise CPU 210 .
- various I/O devices such as a disk controller 204 , a graphics adapter 224 , a video controller 202 , a keyboard adapter 226 , a mouse adapter 206 , a network adapter 220 , and other I/O devices 222 can be coupled to system bus 214 .
- Keyboard adapter 226 and mouse adapter 206 are coupled to keyboard 104 ( FIGS. 1-2 ) and mouse 110 ( FIGS. 1-2 ), respectively, of computer system 100 ( FIG. 1 ).
- graphics adapter 224 and video controller 202 are indicated as distinct units in FIG. 2 , video controller 202 can be integrated into graphics adapter 224 , or vice versa in other embodiments.
- Video controller 202 is suitable for refreshing monitor 106 ( FIGS. 1-2 ) to display images on a screen 108 ( FIG. 1 ) of computer system 100 ( FIG. 1 ).
- Disk controller 204 can control hard drive 114 ( FIGS. 1-2 ), USB port 112 ( FIGS. 1-2 ), and CD-ROM drive 116 ( FIGS. 1-2 ). In other embodiments, distinct units can be used to control each of these devices separately.
- network adapter 220 can comprise and/or be implemented as a WNIC (wireless network interface controller) card (not shown) plugged or coupled to an expansion port (not shown) in computer system 100 ( FIG. 1 ).
- the WNIC card can be a wireless network card built into computer system 100 ( FIG. 1 ).
- a wireless network adapter can be built into computer system 100 by having wireless communication capabilities integrated into the motherboard chipset (not shown), or implemented via one or more dedicated wireless communication chips (not shown), connected through a PCI (peripheral component interconnector) or a PCI express bus of computer system 100 ( FIG. 1 ) or USB port 112 ( FIG. 1 ).
- network adapter 220 can comprise and/or be implemented as a wired network interface controller card (not shown).
- FIG. 1 Although many other components of computer system 100 ( FIG. 1 ) are not shown, such components and their interconnection are well known to those of ordinary skill in the art. Accordingly, further details concerning the construction and composition of computer system 100 and the circuit boards inside chassis 102 ( FIG. 1 ) are not discussed herein.
- computer system 100 is illustrated as a desktop computer in FIG. 1 , there can be examples where computer system 100 can take a different form factor while still having functional elements similar to those described for computer system 100 .
- computer system 100 can comprise a single computer, a single server, or a cluster or collection of computers or servers, or a cloud of computers or servers. Typically, a cluster or collection of servers can be used when the demand on computer system 100 exceeds the reasonable capability of a single server or computer.
- computer system 100 can comprise a portable computer, such as a laptop computer.
- computer system 100 can comprise a mobile device, such as a smart phone or a tablet.
- computer system 100 can comprise an embedded system.
- computer system 100 might or might not contain each of the items shown in FIG. 1 or 2 or might contain multiple of each of the items shown in FIG. 1 or 2 .
- certain implementations of computer system 100 might not contain a CD-ROM, DVD, or Blu-Ray drive 116 .
- Other implementations of computer system 100 might contain two CD-ROM, DVD, or Blu-Ray drives 116 .
- Other implementations of computer system 100 can contain 2 or more monitors 106 .
- Other implementations of computer system 100 could contain no monitors.
- Other implementations of computer system 100 can contain equivalents to certain items.
- hard drive 114 can be replaced or augmented by a solid-state drive (SSD).
- the Internet gives users the ability to research any product or service that the user wishes to purchase to a much greater extent than possible before the Internet. For example, users can go to an eCommerce website, search for a product or service, and analyze the opinions of many different people regarding the product or service. This ability can be particularly useful for users who are comparing two competing products or services.
- URC user generated content
- ratings might not accurately reflect a user's actual views of a product.
- a user might give a low rating for a product, but the low rating is solely due to a slower delivery than expected (thus becoming a review of the eCommerce vendor as opposed to the product).
- Another review might give a low rating on a product for reasons that might not be important to other users. For example, one review of a smartphone could give a low rating for a particular smartphone primarily because of poor camera capabilities. But another user might not care about camera capabilities because she is not interested in photos.
- Another example might involve a more thorough review that analyzes several different aspects of a product and assigns an overall rating of the product.
- some users might be particularly interested in some aspects of the review, but not other aspects of the review.
- some users might not care how comfortable the back seat of a car is because they do not plan to use the back seat on a regular basis. So a review of a car that is otherwise positive, but gives an overall neutral or negative review to a car because of a small back seat, might not be relevant to some users.
- Sentiment analysis can be used to provide an automated method to analyze reviews.
- sentiment analysis is mostly limited to the overall sentiment projected by the entire text. Therefore, for any product, there is no existing automated system that groups features of the product with the sentiment expressed specifically with regard to that feature.
- the representations that come closest to this ideal are the review systems of many e-commerce sites that provide overall ratings for a product based on entire reviews given by the user. These overall ratings usually range from one to five stars for the product, and these overall ratings are explicitly added by the reviewer.
- overall ratings usually range from one to five stars for the product, and these overall ratings are explicitly added by the reviewer.
- One solution is to provide the user with not only an overall rating for the product, but also the underlying features along with a corresponding score for each feature.
- an embodiment can apply sentiment analysis to the text to arrive at a score for each of its features.
- Some embodiments have the capability to concentrate on only the relevant content from a review for every such feature and derive a rating for it. These embodiments provide a better representation and a deeper insight to the product from the data available.
- an embodiment can be more informative than an overall one-star to five-star rating system that is prevalent in today's rating systems.
- FIG. 3 is a flowchart illustrating a method 300 for calculating a feature score.
- the search being performed can be performed on a computer system 100 ( FIG. 1 ).
- Method 300 is merely exemplary and is not limited to the embodiments presented herein. Method 300 can be employed in many different embodiments or examples not specifically depicted or described herein.
- the procedures, the processes and/or the activities of method 300 can be performed in the order presented. In other embodiments, the procedures, the processes, and/or the activities of method 300 can be performed in any other suitable order. In still other embodiments, one or more of the procedures, the processes, and/or the activities of method 300 can be combined or skipped.
- a feature set for the product is produced (block 302 ).
- the feature set will be the set of attributes for which the analysis will be done. There can be several manners in which a feature set can be created.
- the specifications can be analyzed, and the most important attributes mentioned in its specifications can be added to the feature set.
- the feature set of a competing product can be analyzed to produce a feature set for the product being analyzed.
- other methods can be used such as, for example, a single scan of all the reviews can be executed to obtain a list of relevant features on the basis of term-frequency.
- synonyms and related words for each feature can be added to the feature set. It should be understood that this technique is not limited to synonyms. For example, some reviews might refer to the “weight” of a product while others might refer to the “mass” of a product. Related words can be more encompassing than just synonyms. Referring to the same example of weight/mass, some reviews might refer to a product as “light,” and other reviews might refer to the same product as “heavy,” sometimes without even using the terms “weight” or “mass.” The use of modifiers also can be taken into account. For example, the phrases “not heavy” and “very heavy” should be analyzed very differently. Terms such as the above can be added to the feature set as terms to be analyzed.
- An exemplary feature set for a camera might be (body, lens, zoom, flash, screen battery, focus, weight, face-detection, pixels, etc.)
- the user-generated content (“UGC”) is analyzed to find mentions of each feature (block 304 ). Thereafter, each of the mentions can be analyzed to determine polarity of each mention (block 306 ). Any known semantic analysis tool can be used to perform this analysis. In one embodiment, Alchemy API can be used to perform this analysis. Normalizing the total polarity in terms of number of occurrences provides the feature score for the feature in that review. Additionally, the number of occurrences can be an indication of the importance of the feature in the review.
- the number of occurrences in review R j is n ij .
- the sentences that will be considered are the ones that are underlined above.
- a sentiment analysis for each of these sentences is performed.
- the polarity can be recorded as s 1 , s 2 , and s 3 .
- a sentiment analysis system can assign each of these polarities between ⁇ 1 and 1, depending on the content of the sentence.
- the scale used for polarities can range between any two numbers.
- the polarity can be between ⁇ 1 and 1.
- the polarity can be between 0 and 10.
- the polarity can be between 0 and 100.
- An embodiment can readily convert from one polarity scale to a different polarity scale through the use of addition, subtraction, multiplication, division, and the like.
- a table 400 is presented that illustrates an exemplary case with multiple reviews and features.
- Each of columns 410 , 420 , and 430 represent different reviews.
- Each of rows 415 , 425 , 435 , and 445 represents different features of the product being reviewed.
- the intersection of each row and column represents each feature in each review.
- At the intersection of each row and column is one or more polarities s i .
- the table presented in FIG. 4 is merely exemplary. In actual use, there will likely be many more than three reviews and more than four features. In addition, a similar table can exist for each product in a particular database. Some databases contain thousands or even millions of products (for example, the databases of large eCommerce providers).
- the scores over all occurrences of a feature are normalized to obtain feature score Su (block 308 ).
- An exemplary formula to calculate a feature score is as follows:
- a table 500 is presented that illustrates an exemplary case with multiple reviews and features.
- the table presented in FIG. 5 is merely exemplary. In actual use, there will likely be many more than three reviews and many more than four features.
- a similar table can exist for each product in a particular database. Some databases contain thousands or even millions of products (for example, the databases of large eCommerce providers).
- Each of columns 510 , 520 , and 530 represent different reviews.
- Each of rows 515 , 525 , 535 , and 545 represents different features of the product being reviewed.
- the intersection of each row and column represents each feature in each review.
- the feature score Su is calculated in a manner such as that presented above.
- FIG. 5 is a different representation of the data from FIG. 4 .
- feature 1 of review 1 had three different polarities.
- the three different polarities of feature 1 of review 1 have been combined to form a feature score for that review.
- An embodiment can access table 500 , aggregate all the scores for a feature, and normalize the scores by the number of reviews to obtain the final feature score.
- the aggregation of scores can be represented by block 310 .
- the feature score can be displayed to a user (block 312 ).
- an embodiment can be configured to receive an order that purchases a product. Thereafter, an embodiment can be configured to facilitate the delivery of the product to the user. Such a facilitation can be accomplished in one of a variety of ways known in the art or developed in the future.
- the information from the order can be sent to a fulfillment center which will perform delivery services.
- the information can be sent to a brick and mortar store, which can place the product in a special area for pick-up at a later time by the user.
- each of the feature scores can be averaged across every review or across every mention in every review.
- An exemplary result of this technique is shown as table 600 in FIG. 6 .
- column 610 is the only column and represents the aggregate feature score.
- Rows 615 , 625 , 635 , and 645 are each of the features.
- the table presented in FIG. 6 is merely exemplary. In actual use, there can be more than four features. In addition, a similar table can exist for each product in a particular database. Some databases contain thousands or even millions of products (for example, the databases of large eCommerce providers).
- an embodiment now has a score of every feature for each review. If the score s ij is above a threshold (e.g., zero), then the feature f i can be determined to be a “good” feature based on the review R j . Similarly, if the score is below a threshold (e.g., 0), then the feature f i can be determined to be a “bad” feature based on the review R j .
- the threshold between a “good” feature and a “bad” feature can be different from that described above. In an embodiment where the threshold is 0 for both good features and bad features, the feature score can be only good or be bad (except for the unusual case of a feature score being exactly 0).
- the threshold scores can be set such that there are some features that are neither good nor bad.
- the threshold for being considered a “good” feature can be above 0.5.
- the threshold for being considered a “bad” feature can be negative 0.5.
- the threshold for being considered “good” or “bad” can be adjusted by a user.
- all feature scores can be displayed to a user.
- only “good” feature scores can be displayed to a user.
- only “bad” feature scores can be displayed to a user.
- both “good” and “bad” feature scores can be displayed to a user.
- a first threshold score is established. The feature score for each feature of a product is compared to the first threshold score. Thereafter, only feature scores greater than the first threshold score are displayed.
- a second threshold score can be established. Thereafter, only feature scores lower than the second threshold score are displayed.
- Establishing the first or second threshold score can occur in a number of different manners.
- the first or second threshold score can be pre-determined.
- the first or second threshold scores can be set by a user.
- setting a threshold score can be accomplished by using a drop down box on a web page.
- setting a threshold score can be accomplished by using a filter. Other methods of setting a threshold score can be used.
- FIG. 7 an exemplary screen shot 700 illustrating the presentation of feature scores is presented.
- FIG. 7 is merely exemplary and embodiments of the screen representation and menu system are not limited to the embodiments presented herein.
- the screen representation and menu system can be employed in many different embodiments or examples not specifically depicted or described herein.
- screen shot 700 can be shown on a screen of refreshing monitor 106 ( FIG. 1 ).
- exemplary screen shot 700 is name 702 of the product for which the review scores are applicable. It should be understood that while only the name of the product is shown in screen shot 700 , additional information might also be present in various embodiments. For example, a photo of the product, a brief description of the product, the manufacturer of the product, and other relevant information might be listed. It should also be understood that, while the example here pertains to a camera and various features of a camera, embodiments can be used on products of any type. In addition, embodiments are not limited to products, as embodiments can be used to rate services (such as restaurants, auto repair shops, hair salons, and the like) as well.
- rate services such as restaurants, auto repair shops, hair salons, and the like
- Underneath name 702 is a list of various features.
- features include features 710 , 720 , 730 , 740 , 750 , and 760 .
- Each of the features is listed next to a feature score, such as scores 714 , 724 , 734 , 744 , 754 , and 764 .
- Scores 714 , 724 , 734 , 744 , 754 , and 764 are shown in numerical form in FIG. 7 .
- Some embodiments might also include a graphical representation of the feature score.
- such graphical representations can include bar graphs 717 , 727 , 737 , 747 , 757 , and 767 . It should be understood that the graphical representation is not limited to a bar graph and can take a variety of different formats.
- Screen shot 700 can be displayed to a user at various times and in various manners.
- a user might enter an eCommerce site and type in a search term for a camera.
- the result could include the layout of FIG. 7 for multiple cameras at once.
- a user can easily compare different features among different cameras. For example, in FIG. 7 , flash has a relatively low score of 37. So some users might view that negatively and decide to purchase a different camera with a higher score for flash. But a different set of users might not be interested in flash performance and might be interested by the relatively high body score of 86.
- the feature scores can be hidden, except the overall score.
- the result would be similar to many eCommerce sites today, with only an overall score being presented.
- an option such as a drop box, check box, radio button, and the like
- additional feature scores can be presented to the user.
- an embodiment can serve as a one-stop solution. Such a solution can help in retaining and attracting new customers.
- FIG. 8 illustrates a block diagram of a system 800 arranged to perform various tasks.
- System 800 is merely exemplary and is not limited to the embodiments presented herein.
- System 800 can be employed in many different embodiments or examples not specifically depicted or described herein.
- certain elements or modules of system 800 can perform various procedures, processes, and/or acts.
- the procedures, processes, and/or acts can be performed by other suitable elements or modules.
- system 800 can be implemented using computer system 100 ( FIG. 1 ).
- system 800 can include feature set producing module 802 .
- feature set producing module 802 can perform block 302 ( FIG. 3 ) of producing a feature set.
- system 800 also can include a UGC analysis module 804 .
- UGC analysis module 804 can perform block 304 ( FIG. 3 ) of analyzing UGC to find mentions of each feature.
- system 800 further can include a total polarity determination module 806 .
- total polarity determination module 806 can perform block 306 ( FIG. 3 ) of determining total polarity by analyzing mentions.
- system 800 additionally can include feature score calculation module 808 .
- feature score calculation module 808 can perform block 308 ( FIG. 3 ) of presenting search results to the requester.
- order system 800 additionally can include final feature score calculation module 810 .
- final feature score calculation module 810 can perform block 310 ( FIG. 3 ) of calculating a final feature score.
- order system 800 additionally can include feature score display module 812 .
- feature score display module 812 can perform block 312 ( FIG. 3 ) of displaying feature scores to a user.
- order system 800 additionally can include order receiving module.
- order receiving module can perform receiving an order from a user.
- order system 800 additionally can include delivery facilitation module 816 .
- delivery facilitation module 816 can perform block 316 ( FIG. 3 ) of facilitating the delivery of the product to a user.
- Another dimension is added to reviews. Annotating the reviews with relevant features of the product makes the reviews more readable for users. Thus, information is more accessible to users as the user can read only the reviews that discuss the features of interest to him/her.
- Another advantage is the ability to use vertical scoring. Instead of there being a single rating for the overall review where the single rating might not accurately reflect what is important to some users, an analysis of features talked about in the review can be obtained. Providing a feature score goes beyond product scoring. Features can be highlighted depending upon good or bad sentiment expressed and feature scores of different products can help in comparing the products using only a single rating.
- This method can be used for any textual content. This approach can be enhanced and tuned to work with other textual contents to present sentiment expressed corresponding to various issues related to the content.
- an eCommerce provider can determine the best and most popular feature of a product from the calculated feature scores and use those features as a selling point or a badge for the product to highlight it.
- the eCommerce provider can sell information for specific products and/or general product categories to product designers and manufacturers.
- the feature score can be used to annotate reviews for which the feature score is not equal to zero. If a feature has either a positive feature score or a negative feature score, the implication is that the review mentions that particular feature. It is also possible for a review to have a neutral feature score even if the feature is mentioned in the review. For example, a review can mention a certain feature twice, once with a positive connotation and once with a negative connotation. There are a variety of different ways in which such a situation can be handled. In some embodiments, all feature scores that are zero can be ignored. In some embodiments, all mentions of a feature can be displayed. In such a manner, if a particular review mentions a feature, that mention can be indicated even if the feature score for that feature is zero.
- the scale used to determine if a pair is “Bad” or “Good” can range between any two numbers.
- the feature score can range between ⁇ 1 and 1.
- the feature score can range between 0 and 10.
- the feature score can range between 1 and 100.
- An embodiment can readily convert from one polarity scale to a different polarity scale through the use of addition, subtraction, multiplication, division, and the like.
- the threshold score being used to determine if a score is “Bad” or “Good” can be changed depending a user's desires.
- the threshold score can be changed by a user on a website.
- the threshold score is not changeable by a typical user.
- a score can be considered “good” if the score is over 0, while a score can be considered “bad” if the score is 0 or below.
- a score can be considered “good” if the score is over 0.2, while a score of ⁇ 0.2 or less can be considered “bad,” with scores between ⁇ 0.2 and 0.2 not being considered “good” or “bad.”
- Other scores could also be used as a threshold score for determining if a review is “good,” “bad,” or “neutral.”
- Method 900 is merely exemplary and is not limited to the embodiments presented herein. Method 900 can be employed in many different embodiments or examples not specifically depicted or described herein. In some embodiments, the procedures, the processes and/or the activities of method 900 can be performed in the order presented. In other embodiments, the procedures, the processes, and/or the activities of method 900 can be performed in any other suitable order. In still other embodiments, one or more of the procedures, the processes, and/or the activities of method 900 can be combined or skipped.
- each review after feature scores are obtained (using, for example, a method described above), one would tabulate each review and each feature score in that review (block 902 ).
- Each of the m features can have a feature score, established in a manner such as that set forth above.
- the reviews and feature sets can be combined to form a two-dimensional matrix (block 904 ): ⁇ S 11 , S 12 , . . . , S 1N ; S 21 , S 22 , . . . , S 2N ; S M1 , S M2 , . . . , S MN ⁇ .
- the reviews are associated with features (block 906 ).
- review R 1 has features f 1 , f 2 , and f 4 that have feature scores that are not equal to zero.
- R 1 gets associated with f 1 , f 2 , and f 4 .
- each review is matched with the features that are discussed in the review.
- the features are labeled depending on the feature score (block 908 ).
- Features with a feature score greater than a given threshold score e.g., zero
- the features with a feature score less than a given threshold score e.g., zero
- the reviews can be displayed to the user with a tag indicating the feature and whether the feature score was “Good” or “Bad” (block 910 ). Displaying a review to the user can involve transmitting data to a user's computer that cause the user's computer to display reviews.
- the above-presented annotation of reviews can help make the reviews more readable to a user.
- an embodiment can lead to a display of reviews that is more informative. A user is able to focus only on reviews that concern features of interest and ignore reviews that concern features with which he is not concerned.
- Browsing customer reviews and other UGC can be as simple as viewing annotations of each review that note the attributes and features that each review discusses. Also, the sentiment of the author can be highlighted.
- an embodiment annotates reviews with the features that each reviews discusses. In such a manner, a user can quickly determine what each review is discussing and what each review is about without the need to read the entire review.
- FIG. 10 is merely exemplary and embodiments of the screen representation are not limited to the embodiments presented herein.
- the screen representation can be employed in many different embodiments or examples not specifically depicted or described herein.
- FIG. 10 there happens to be four reviews. This is merely exemplary—there can be many more than four reviews present on a single screen, or there can be fewer than four reviews on a single screen.
- Column 1010 contains a description of each review.
- the description is entered by a user.
- a user submitting a review will often submit a title of the review.
- the title of the review is not pertinent to actual content of the review, making the title of the review not as useful as it could be.
- review 1025 is titled, “disappointed.” Such a title informs a user that the reviewer was not satisfied for some reason, but the user is left to guess at the reason.
- review 1045 is titled, “slow shipping.” Such a title tells the user nothing about the product except that the delivery was slower than expected.
- Column 1020 contains a list of “tags” that represent features that received either positive or negative feature scores in each review.
- the tags can be color-coded. In such a manner, one color can represent positive feature scores and another color can represent negative feature scores.
- color coding some embodiments can use other methods of indicating the difference between positive, negative, and neutral feature scores, such as shading, hatch marks, underlining, typeface, font colors, and the like.
- a user is able to click on a tag to view only those reviews that contain features scores for that feature.
- FIG. 11 a screen shot 1100 is presented.
- FIG. 11 is merely exemplary and embodiments of the screen representation are not limited to the embodiments presented herein.
- the screen representation can be employed in many different embodiments or examples not specifically depicted or described herein.
- Screen 1100 illustrates a result of a user clicking on the “body” tag from screen 1000 .
- Screen 1100 illustrates a result of a user clicking on the “body” tag from screen 1000 .
- only reviews that have a non-zero feature score for the selected feature are presented to the user. It should be understood that it is also possible to present products with a feature score of zero to the user.
- the title 1170 can serve to remind the user as to which product and feature the reviews pertain. Such a feature can be handy because a user might have several browser windows or tabs open at once; if switching back and forth between reviews of different products, it can be useful to be reminded what the product and features the user is viewing.
- Reviews 1115 , 1125 , and 1135 are presented to the user. On a large eCommerce site, the most popular products can have thousands of reviews. Only a portion of the reviews might be visible at a time. In the illustration of FIG. 11 , there happens to be three reviews. It should be understood that there can be many more than three reviews present on a single screen.
- the reviews can contain two columns, column 1110 and column 1120 .
- Column 1110 displays the title of the review.
- only the tag that is the subject of the search is displayed to the user in column 1120 .
- all of the pertinent tags can be displayed to the user, to allow the user to view other feature scores.
- an indication of whether the review is positive or negative can be contained in column 1120 .
- the user is able to select on a review (such as by clicking on the review title in column 1110 ) to view an excerpt of the review, ranging up to the entire review.
- a user might wish to read an excerpt or the entire review to judge how reliable the reviewer is.
- the text of the review, or an excerpt thereof is displayed to the user.
- there can be elements of the review that can be highlighted, underlined, or otherwise distinguished from the remainder of the review.
- the sentences of the review that contain a reference to the “body” feature can be highlighted, underlined, or otherwise distinguished. The methods used to perform such a feature are detailed below.
- a graphical indicia that represents the distribution of positive and negative features from the reviews.
- Pictured in FIG. 11 is a pie chart 1150 .
- Pie chart 1150 can contain various segments, such as positive segment 1160 , neutral segment 1162 , and negative segment 1164 . Each of these segments is configured to illustrate the percentage of reviews that contain a certain feature score.
- the vast majority of reviews that contain a feature score for “body” are positive, so positive segment 1160 is much larger than the other two segments.
- positive segment 1160 can be a first color
- neutral segment 1162 can be a second color
- negative segment 1164 can be a third color.
- the color-coding can extend to the tags in column 1120 .
- a review that contained a positive feature score for “body” can be the same first color as positive segment 1160 .
- a review that contained a neutral feature score for “body” can be the same second color as neutral segment 1162 and a review that contained a negative feature score for “body” can be the same third color as negative segment 1164 .
- a pie chart is illustrated in FIG. 11
- other embodiments can use other types of graphical indicia, such as bar graphs, line graphs, histograms, and the like.
- the creation of the graphical indicia can be accomplished in a variety of different manners.
- the number of reviews featuring a particular feature is known or can be determined through common database operations.
- the creation of a graphical representation can be accomplished by determining the total number of reviews with a particular feature, the number of positive reviews of that particular feature, the number of negative reviews of that particular feature, and the number of neutral reviews of that particular features.
- graphical indicia can be created.
- the number of reviews that discuss a particular feature can be compared to the total number of reviews of the product. In such a manner, a user can more easily determine how important a feature is to the reviews.
- an embodiment would provide a solution that would help in retaining old customers and attracting new customers.
- An embodiment can also lead to a customer spending more time on a particular site, which can lead to increased revenue generation for an eCommerce provider using an embodiment.
- Products can have many features that a prospective purchaser might have to take into consideration before making a purchase.
- a digital single lens reflex (DSLR) camera might have more than 50 features to consider before buying it.
- a person shopping for a DSLR might not know exactly what he needs and what features he should concentrate on. Even after knowing the features you should look into, it can still be desirable to read the reviews.
- the reviews can give an insight as to how well the features work, where the product excels, and where the product needs improvement.
- the filtering procedure can use heuristics to prioritize features based on popularity and bias of the features. If a feature if highly regarded, then its popularity increases. If the sentiment of a sentence is highly positive or highly negative, the sentence can be considered to be highly biased.
- the range of feature scores is from 0 to 1. In other embodiments, a different range of scores can be used. In an embodiment where the range of feature scores is from 0 to 1, the feature ranking can be calculated as follows:
- Method 1300 is merely exemplary and is not limited to the embodiments presented herein. Method 1300 can be employed in many different embodiments or examples not specifically depicted or described herein. In some embodiments, the procedures, the processes and/or the activities of method 1300 can be performed in the order presented. In other embodiments, the procedures, the processes, and/or the activities of method 1300 can be performed in any other suitable order. In still other embodiments, one or more of the procedures, the processes, and/or the activities of method 1300 can be combined or skipped.
- the normalized score ranges between 0 and 1.
- Feature scores are generated for every feature of a product, possibly by using a method described earlier in this specification (block 1302 ). In general, this process would involve analyzing each review, determining the features described in each review, and assigning a feature score to each feature in each review. Once all the reviews are analyzed, all of the features with feature scores are determined (block 1304 ). The features can be sorted in descending order of feature score (block 1306 ). Thereafter, the top K features can be selected (block 1308 ). K can be selected to be any number. As described above, some products can have more than 50 features. It might not be desirable to list all the features of a product, as the list can be too long.
- the top K features can be restricted by setting a minimum threshold score T (block 1310 ). For example, because the normalized feature scores range from 0 (bad) to 1 (good), a threshold score of 0.5 can be set. Thereafter, if only 5 features have a feature score of 0.5, only those 5 features are displayed to the user. It should be understood that the values of K and T can be fine-tuned depending on the desired results.
- a visual indicia of the top features can be created (block 1312 ).
- the visual indicia can be a tag cloud.
- a tag cloud is a visual representation of data, in this case, of features. The importance of each feature is shown by various features, such as font size and color.
- An exemplary tag cloud is shown in FIG. 14 .
- FIG. 14 is merely exemplary and embodiments of the screen representation are not limited to the embodiments presented herein.
- the screen representation can be employed in many different embodiments or examples not specifically depicted or described herein.
- Screen shot 1400 displays a portion of an exemplary display.
- the data used to create screen shot 1400 can be transmitted to a user's computer, tablet, smartphone, or other similar device along with instructions to the user's computer, tablet, smartphone, or other similar device that can be arranged to cause the device to display screen shot 1400 .
- Tag cloud 1450 contains tag cloud 1450 .
- tag cloud 1450 includes each of the K features of an exemplary camera that have a feature score over threshold T.
- tag cloud 1450 displays each of the features with a different font size to differentiate between the features.
- the features of tag cloud 1450 are distinguished by the feature score of each feature—the higher the feature score, the larger the font size of the feature.
- Tag cloud 1450 contains 12 different features: face detection, pixel, battery, LCD (liquid crystal display), photo, screen, zoom, video, body, focus, lens, flash, and display.
- the features are sized and/or colored to visually illustrate the score of the feature. For example, the feature “zoom” has the largest font size, and therefore has the highest feature score of any of the features. In contrast, the feature “lcd” has the smallest font size, and therefore has the lowest feature score over threshold T.
- the size of the font can instead be used to illustrate the number of reviews with that feature and the color of the font can be used to illustrate the feature score.
- the color of the font can be used to illustrate the feature score.
- a legend can be provided to aid a viewer in determining that, for example, the more green a font color is, the more higher the feature score, while the more red a font color is, the lower the feature score is.
- Method 1500 can be performed on a computer system such as computer system 100 ( FIG. 1 ).
- Method 1500 is merely exemplary and is not limited to the embodiments presented herein.
- Method 1500 can be employed in many different embodiments or examples not specifically depicted or described herein.
- the procedures, the processes and/or the activities of method 1500 can be performed in the order presented.
- the procedures, the processes, and/or the activities of method 1500 can be performed in any other suitable order.
- one or more of the procedures, the processes, and/or the activities of method 1500 can be combined or skipped.
- the elements of method 1500 can be performed after block 1312 of FIG. 13 . However, the elements of method 1500 can also be performed independently or as a part of a different method.
- the top K features and their feature scores are sorted by feature score (block 1502 ).
- the tag cloud can be based on various aspects of reviews. The embodiment illustrated in FIG. 15 will use feature score. However, other embodiments, can use other aspects, such as number of reviews.
- the top K features can be determined in one of a variety of different manners. Assuming the polarity scores range from 0 (for lowest rating) to 1 (for the highest rating), in one embodiment, the sum of the polarity scores can be used to create an unnormalized score. In some embodiments, this value can be the sum of the maximum of the polarity and 1 minus the polarity, such that the lowest reviews of a feature are captured along with the highest reviews of a feature. In other words, a user might not be interested in just positive reviews of a certain feature. The user also wants to know what people score a feature in a negative manner. The polarity ranges from 0 to 1. Using both the polarity score and 1—the polarity score ensures that a feature score of 0.8 is treated in the same manner as a feature score of 0.2.
- each unnormalized score can be determined by taking the unnoramlized score, subtracting the smallest unnormalized score, and dividing the result by the highest unnormalized score.
- the largest font size F is determined (block 1504 ).
- the largest font size is dependent on the medium by which the font will be displayed. For example, an embodiment being displayed on a mobile device might have a different largest font size than an embodiment being displayed on a desktop device.
- the font size of each of the features is determined (block 1506 ).
- the font size can be distributed linearly. In some embodiments, the font size can be distributed logarithmically.
- the feature with the highest feature score is chosen to have the largest font size F.
- the feature with the second highest feature score is chosen to have the font size (K ⁇ 1/K)*F.
- the feature with the third highest feature score is chosen to have the font size (K ⁇ 2/K)*F. And so on until the feature with the Kth lowest feature score has the font size (1/K)*F.
- the font size is chosen by proportionally scaling the “normalized feature score” between a range of font sizes between a minimum font size and a maximum font size.
- the minimum font size is approximately 5 points and the maximum font size is approximately 72 points. Other font sizes can be used in other embodiments.
- the features are arranged in a random order (block 1508 ). Thereafter, the random arrangement of features in different font sizes is transmitted to a user for display (block 1510 ).
- the tag cloud can be configured to be clickable.
- a user can click on one of the features listed in the tag cloud and be shown a list of reviews featuring that feature (such as the bottom portion of screen shot 1100 of FIG. 11 ).
- the tag cloud can be made clickable in one of a variety of ways known in the art. For example, an image map can be created based on the random arrangement of features and font sizes. Thereafter, the image map can be laid over the tag cloud in such a manner that clicking on an area of the tag cloud results in clicking an area of the image map corresponding to the clicked area of the tag cloud. For example, with respect to FIG. 14 , when a user clicks on the area corresponding to “zoom,” the coordinates of the area being clicked are s corresponding area of the image map.
- the image map has a set of coordinates that correspond to each feature.
- an embodiment can help a user to understand a special aspect of a set of reviews.
- the content can be annotated with sentiment and highlights. If the user is interested in acquiring knowledge about a specific aspect from multiple contents, a visualization of the content can be a great help to the user.
- the sentiment of the content also the sentiment related to the specific aspect.
- pie chart 1150 shows the distribution of different sentiment in the reviews. Different reviews along with their sentiment are listed. On clicking a specific review, an annotated version of the full review expands. The result is shown in FIG. 12 .
- the “Zoom” feature has positive, negative, and neutral reviews. These reviews can be highlighted in different colors to indicate the different categories. Such highlighting can help the user determine the sentences of the reviews that he can concentrate upon. It should be understood that the highlighting can use three different colors, three different types of underlining, or any types of techniques that can be used to distinguish between differences in reviews. In addition, other numbers of colors, underlining, or highlighting can be used. In some embodiments, only two different colors, underlining, or highlighting can be used, with neutral reviews being uncolored. In some embodiments, only one color, underlining, or highlighting is used. In some embodiments, a user can select how many different color, underlining, or highlighting is used and can select which sentences (good, bad, or neutral) get colored, underlined, or highlighted.
- Method 1600 is merely exemplary and is not limited to the embodiments presented herein. Method 1600 can be employed in many different embodiments or examples not specifically depicted or described herein. In some embodiments, the procedures, the processes and/or the activities of method 1600 can be performed in the order presented. In other embodiments, the procedures, the processes, and/or the activities of method 1600 can be performed in any other suitable order. In still other embodiments, one or more of the procedures, the processes, and/or the activities of method 1600 can be combined or skipped.
- Each review is analyzed and a feature score is generated for every feature of a product, possibly using a method described earlier in this specification.
- Feature scores are generated for every feature of a product, possibly by using a method described earlier in this specification (block 1602 ).
- this process would involve analyzing each review, determining the features described in each review, and assigning a feature score to each feature in each review.
- each portion of the review that mentions a feature is noted, such that every mention of every feature has a notation in a database (block 1604 ). For example, a review that mentions both the features “body” and “zoom” would be marked with both of those features.
- the sentence or sentences in each review that mention “body” or “zoom” would be noted.
- an embodiment can receive input from a user indicating a desire to view details about a certain feature (block 1606 ).
- a graphical indicia of the reviews can be displayed to the user, along with one or more reviews, as illustrated above with respect to FIG. 11 .
- a pie chart or other graphical indicia can be used to illustrate the distribution between positive reviews, negative reviews, and neutral reviews.
- a summary sentence is displayed to the user, along with an indication of whether the review was positive or negative.
- the user can then click on one of the reviews, resulting in the summary shown in FIG. 12 .
- element 1240 The entire review is now shown to the user as element 1240 .
- element 1242 shows a sentence discussing the feature in a negative manner.
- element 1244 shows a sentence discussing the same feature in a neutral manner.
- Element 1246 shows a sentence discussing the same feature in a positive manner.
- the creation of such a display can be discussed.
- the database containing the reviews is consulted.
- the score of each mention of each feature is accessed to determine if the mention is positive, negative, or neutral. Thereafter, the mention is displayed in a different manner depending on the mention (block 1608 ).
- the review portion would be stored in a single field. Thus, even a 1000-word (or longer) review would be contained in a single field.
- the review can be parsed, then stored in a plurality of fields in a database. For example, each sentence can be stored in a separate field. Also stored in a separate field is the feature that the sentence discusses. In such a manner, as each sentence of a review is displayed, special formatting can be applied to each sentence, depending on the content of the sentence. When the text of a review is to be displayed, each sentence can be retrieved individually. Then, the field indicating which feature is discussed by the sentence is accessed to determine if the sentence should be emphasized in some way.
- embodiments and limitations disclosed herein are not dedicated to the public under the doctrine of dedication if the embodiments and/or limitations: (1) are not expressly claimed in the claims; and (2) are or are potentially equivalents of express elements and/or limitations in the claims under the doctrine of equivalents.
Abstract
Description
Total polarity (f i ,R j)=Σk=1 n
Claims (23)
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US14/846,501 US10140646B2 (en) | 2015-09-04 | 2015-09-04 | System and method for analyzing features in product reviews and displaying the results |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US14/846,501 US10140646B2 (en) | 2015-09-04 | 2015-09-04 | System and method for analyzing features in product reviews and displaying the results |
Publications (2)
Publication Number | Publication Date |
---|---|
US20170068648A1 US20170068648A1 (en) | 2017-03-09 |
US10140646B2 true US10140646B2 (en) | 2018-11-27 |
Family
ID=58190015
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US14/846,501 Active 2035-10-04 US10140646B2 (en) | 2015-09-04 | 2015-09-04 | System and method for analyzing features in product reviews and displaying the results |
Country Status (1)
Country | Link |
---|---|
US (1) | US10140646B2 (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US11107092B2 (en) * | 2019-01-18 | 2021-08-31 | Sprinklr, Inc. | Content insight system |
US11144730B2 (en) | 2019-08-08 | 2021-10-12 | Sprinklr, Inc. | Modeling end to end dialogues using intent oriented decoding |
US11715134B2 (en) | 2019-06-04 | 2023-08-01 | Sprinklr, Inc. | Content compliance system |
Families Citing this family (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20170178162A1 (en) * | 2015-12-22 | 2017-06-22 | Adobe Systems Incorporated | Control of which Sources are used to Answer Questions in a Digital Medium Environment |
US20190156826A1 (en) * | 2017-11-18 | 2019-05-23 | Cogi, Inc. | Interactive representation of content for relevance detection and review |
EP3884402A4 (en) * | 2018-11-22 | 2022-07-27 | Yeogirl Yun | Multilingual tag-based review system |
CN112597409A (en) * | 2021-03-04 | 2021-04-02 | 蚂蚁智信(杭州)信息技术有限公司 | Label display method and device |
Citations (42)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US4952920A (en) * | 1987-01-20 | 1990-08-28 | Kabushiki Kaisha Toshiba | Display apparatus having horizontal/vertical conversion display functions |
US5371673A (en) * | 1987-04-06 | 1994-12-06 | Fan; David P. | Information processing analysis system for sorting and scoring text |
US20040078214A1 (en) * | 2002-10-21 | 2004-04-22 | Speiser Leonard Robert | Product recommendation in a network-based commerce system |
US20050034071A1 (en) * | 2003-08-08 | 2005-02-10 | Musgrove Timothy A. | System and method for determining quality of written product reviews in an automated manner |
US20050091038A1 (en) * | 2003-10-22 | 2005-04-28 | Jeonghee Yi | Method and system for extracting opinions from text documents |
US20050125216A1 (en) * | 2003-12-05 | 2005-06-09 | Chitrapura Krishna P. | Extracting and grouping opinions from text documents |
US7028250B2 (en) * | 2000-05-25 | 2006-04-11 | Kanisa, Inc. | System and method for automatically classifying text |
US20060129446A1 (en) * | 2004-12-14 | 2006-06-15 | Ruhl Jan M | Method and system for finding and aggregating reviews for a product |
US20060200342A1 (en) * | 2005-03-01 | 2006-09-07 | Microsoft Corporation | System for processing sentiment-bearing text |
US20070078670A1 (en) * | 2005-09-30 | 2007-04-05 | Dave Kushal B | Selecting high quality reviews for display |
US20070078671A1 (en) * | 2005-09-30 | 2007-04-05 | Dave Kushal B | Selecting high quality text within identified reviews for display in review snippets |
US20070244888A1 (en) | 2005-10-03 | 2007-10-18 | Powerreviews, Inc. | Affinity attributes for product assessment |
US20080133488A1 (en) * | 2006-11-22 | 2008-06-05 | Nagaraju Bandaru | Method and system for analyzing user-generated content |
US20080215571A1 (en) * | 2007-03-01 | 2008-09-04 | Microsoft Corporation | Product review search |
US20080249764A1 (en) * | 2007-03-01 | 2008-10-09 | Microsoft Corporation | Smart Sentiment Classifier for Product Reviews |
US20080270116A1 (en) * | 2007-04-24 | 2008-10-30 | Namrata Godbole | Large-Scale Sentiment Analysis |
US7558769B2 (en) * | 2005-09-30 | 2009-07-07 | Google Inc. | Identifying clusters of similar reviews and displaying representative reviews from multiple clusters |
US20090182727A1 (en) * | 2008-01-16 | 2009-07-16 | International Business Machines Corporation | System and method for generating tag cloud in user collaboration websites |
US20090265332A1 (en) * | 2008-04-18 | 2009-10-22 | Biz360 Inc. | System and Methods for Evaluating Feature Opinions for Products, Services, and Entities |
US7620651B2 (en) | 2005-11-15 | 2009-11-17 | Powerreviews, Inc. | System for dynamic product summary based on consumer-contributed keywords |
US20090287989A1 (en) * | 2008-05-16 | 2009-11-19 | International Business Machines Corporation | Presenting tag clouds based on user information |
US20090319342A1 (en) * | 2008-06-19 | 2009-12-24 | Wize, Inc. | System and method for aggregating and summarizing product/topic sentiment |
US20100023311A1 (en) * | 2006-09-13 | 2010-01-28 | Venkatramanan Siva Subrahmanian | System and method for analysis of an opinion expressed in documents with regard to a particular topic |
US20100169317A1 (en) | 2008-12-31 | 2010-07-01 | Microsoft Corporation | Product or Service Review Summarization Using Attributes |
US20110153451A1 (en) * | 2009-12-23 | 2011-06-23 | Sears Brands, Llc | Systems and methods for using a social network to provide product related information |
US8032471B2 (en) | 2007-12-21 | 2011-10-04 | Yahoo! Inc. | System and method for annotating and analyzing reviews with inferred analytics |
US20120110438A1 (en) * | 2010-11-03 | 2012-05-03 | Microsoft Corporation | Proportional Font Scaling |
US20120209751A1 (en) * | 2011-02-11 | 2012-08-16 | Fuji Xerox Co., Ltd. | Systems and methods of generating use-based product searching |
US20120254060A1 (en) * | 2011-04-04 | 2012-10-04 | Northwestern University | System, Method, And Computer Readable Medium for Ranking Products And Services Based On User Reviews |
US8554701B1 (en) | 2011-03-18 | 2013-10-08 | Amazon Technologies, Inc. | Determining sentiment of sentences from customer reviews |
US20130268262A1 (en) * | 2012-04-10 | 2013-10-10 | Theysay Limited | System and Method for Analysing Natural Language |
US8600796B1 (en) * | 2012-01-30 | 2013-12-03 | Bazaarvoice, Inc. | System, method and computer program product for identifying products associated with polarized sentiments |
US8630843B2 (en) | 2011-04-29 | 2014-01-14 | International Business Machines Corporation | Generating snippet for review on the internet |
US8645295B1 (en) | 2009-07-27 | 2014-02-04 | Amazon Technologies, Inc. | Methods and system of associating reviewable attributes with items |
US8700480B1 (en) | 2011-06-20 | 2014-04-15 | Amazon Technologies, Inc. | Extracting quotes from customer reviews regarding collections of items |
US8818788B1 (en) * | 2012-02-01 | 2014-08-26 | Bazaarvoice, Inc. | System, method and computer program product for identifying words within collection of text applicable to specific sentiment |
US20150066803A1 (en) * | 2013-08-27 | 2015-03-05 | International Business Machines Corporation | Quantitative product feature analysis |
US20150286710A1 (en) * | 2014-04-03 | 2015-10-08 | Adobe Systems Incorporated | Contextualized sentiment text analysis vocabulary generation |
US20150324342A1 (en) * | 2013-01-11 | 2015-11-12 | Nokia Technologies Oy | Method and apparatus for enriching social media to improve personalized user experience |
US20170068975A1 (en) * | 2015-09-04 | 2017-03-09 | Wal-Mart Stores, Inc. | System and method for displaying reviews according to features |
US20170124575A1 (en) * | 2015-11-03 | 2017-05-04 | International Business Machines Corporation | Analyzing sentiment in product reviews |
US20170200205A1 (en) * | 2016-01-11 | 2017-07-13 | Medallia, Inc. | Method and system for analyzing user reviews |
-
2015
- 2015-09-04 US US14/846,501 patent/US10140646B2/en active Active
Patent Citations (46)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US4952920A (en) * | 1987-01-20 | 1990-08-28 | Kabushiki Kaisha Toshiba | Display apparatus having horizontal/vertical conversion display functions |
US5371673A (en) * | 1987-04-06 | 1994-12-06 | Fan; David P. | Information processing analysis system for sorting and scoring text |
US7028250B2 (en) * | 2000-05-25 | 2006-04-11 | Kanisa, Inc. | System and method for automatically classifying text |
US20040078214A1 (en) * | 2002-10-21 | 2004-04-22 | Speiser Leonard Robert | Product recommendation in a network-based commerce system |
US20050034071A1 (en) * | 2003-08-08 | 2005-02-10 | Musgrove Timothy A. | System and method for determining quality of written product reviews in an automated manner |
US20050091038A1 (en) * | 2003-10-22 | 2005-04-28 | Jeonghee Yi | Method and system for extracting opinions from text documents |
US20050125216A1 (en) * | 2003-12-05 | 2005-06-09 | Chitrapura Krishna P. | Extracting and grouping opinions from text documents |
US20060129446A1 (en) * | 2004-12-14 | 2006-06-15 | Ruhl Jan M | Method and system for finding and aggregating reviews for a product |
US20060200342A1 (en) * | 2005-03-01 | 2006-09-07 | Microsoft Corporation | System for processing sentiment-bearing text |
US20070078670A1 (en) * | 2005-09-30 | 2007-04-05 | Dave Kushal B | Selecting high quality reviews for display |
US20070078671A1 (en) * | 2005-09-30 | 2007-04-05 | Dave Kushal B | Selecting high quality text within identified reviews for display in review snippets |
US7558769B2 (en) * | 2005-09-30 | 2009-07-07 | Google Inc. | Identifying clusters of similar reviews and displaying representative reviews from multiple clusters |
US20070244888A1 (en) | 2005-10-03 | 2007-10-18 | Powerreviews, Inc. | Affinity attributes for product assessment |
US7937391B2 (en) | 2005-11-15 | 2011-05-03 | Powerreviews, Inc. | Consumer product review system using a comparison chart |
US7620651B2 (en) | 2005-11-15 | 2009-11-17 | Powerreviews, Inc. | System for dynamic product summary based on consumer-contributed keywords |
US20100023311A1 (en) * | 2006-09-13 | 2010-01-28 | Venkatramanan Siva Subrahmanian | System and method for analysis of an opinion expressed in documents with regard to a particular topic |
US7930302B2 (en) | 2006-11-22 | 2011-04-19 | Intuit Inc. | Method and system for analyzing user-generated content |
US20080133488A1 (en) * | 2006-11-22 | 2008-06-05 | Nagaraju Bandaru | Method and system for analyzing user-generated content |
US20080249764A1 (en) * | 2007-03-01 | 2008-10-09 | Microsoft Corporation | Smart Sentiment Classifier for Product Reviews |
US20080215571A1 (en) * | 2007-03-01 | 2008-09-04 | Microsoft Corporation | Product review search |
US20080270116A1 (en) * | 2007-04-24 | 2008-10-30 | Namrata Godbole | Large-Scale Sentiment Analysis |
US8032471B2 (en) | 2007-12-21 | 2011-10-04 | Yahoo! Inc. | System and method for annotating and analyzing reviews with inferred analytics |
US20090182727A1 (en) * | 2008-01-16 | 2009-07-16 | International Business Machines Corporation | System and method for generating tag cloud in user collaboration websites |
US8117207B2 (en) | 2008-04-18 | 2012-02-14 | Biz360 Inc. | System and methods for evaluating feature opinions for products, services, and entities |
US20090265332A1 (en) * | 2008-04-18 | 2009-10-22 | Biz360 Inc. | System and Methods for Evaluating Feature Opinions for Products, Services, and Entities |
US20090287989A1 (en) * | 2008-05-16 | 2009-11-19 | International Business Machines Corporation | Presenting tag clouds based on user information |
US20090319342A1 (en) * | 2008-06-19 | 2009-12-24 | Wize, Inc. | System and method for aggregating and summarizing product/topic sentiment |
US20100169317A1 (en) | 2008-12-31 | 2010-07-01 | Microsoft Corporation | Product or Service Review Summarization Using Attributes |
US8645295B1 (en) | 2009-07-27 | 2014-02-04 | Amazon Technologies, Inc. | Methods and system of associating reviewable attributes with items |
US20110153451A1 (en) * | 2009-12-23 | 2011-06-23 | Sears Brands, Llc | Systems and methods for using a social network to provide product related information |
US20120110438A1 (en) * | 2010-11-03 | 2012-05-03 | Microsoft Corporation | Proportional Font Scaling |
US20120209751A1 (en) * | 2011-02-11 | 2012-08-16 | Fuji Xerox Co., Ltd. | Systems and methods of generating use-based product searching |
US8554701B1 (en) | 2011-03-18 | 2013-10-08 | Amazon Technologies, Inc. | Determining sentiment of sentences from customer reviews |
US20120254060A1 (en) * | 2011-04-04 | 2012-10-04 | Northwestern University | System, Method, And Computer Readable Medium for Ranking Products And Services Based On User Reviews |
US8630843B2 (en) | 2011-04-29 | 2014-01-14 | International Business Machines Corporation | Generating snippet for review on the internet |
US8630845B2 (en) | 2011-04-29 | 2014-01-14 | International Business Machines Corporation | Generating snippet for review on the Internet |
US8700480B1 (en) | 2011-06-20 | 2014-04-15 | Amazon Technologies, Inc. | Extracting quotes from customer reviews regarding collections of items |
US8600796B1 (en) * | 2012-01-30 | 2013-12-03 | Bazaarvoice, Inc. | System, method and computer program product for identifying products associated with polarized sentiments |
US8818788B1 (en) * | 2012-02-01 | 2014-08-26 | Bazaarvoice, Inc. | System, method and computer program product for identifying words within collection of text applicable to specific sentiment |
US20130268262A1 (en) * | 2012-04-10 | 2013-10-10 | Theysay Limited | System and Method for Analysing Natural Language |
US20150324342A1 (en) * | 2013-01-11 | 2015-11-12 | Nokia Technologies Oy | Method and apparatus for enriching social media to improve personalized user experience |
US20150066803A1 (en) * | 2013-08-27 | 2015-03-05 | International Business Machines Corporation | Quantitative product feature analysis |
US20150286710A1 (en) * | 2014-04-03 | 2015-10-08 | Adobe Systems Incorporated | Contextualized sentiment text analysis vocabulary generation |
US20170068975A1 (en) * | 2015-09-04 | 2017-03-09 | Wal-Mart Stores, Inc. | System and method for displaying reviews according to features |
US20170124575A1 (en) * | 2015-11-03 | 2017-05-04 | International Business Machines Corporation | Analyzing sentiment in product reviews |
US20170200205A1 (en) * | 2016-01-11 | 2017-07-13 | Medallia, Inc. | Method and system for analyzing user reviews |
Non-Patent Citations (2)
Title |
---|
Alejandra L. Fernandex, Tony Veale, Prasenjit Majumder, Feature Extraction from Product Reviews using Feature Similarity and Polarity, Technical Report UCD-CSI-2009-09, Oct. 2009. * |
Edgar F. Codd, A Raelational Model of Data for Large Shared Data Banks, vol. 13, Jun. 1970. * |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US11107092B2 (en) * | 2019-01-18 | 2021-08-31 | Sprinklr, Inc. | Content insight system |
US11715134B2 (en) | 2019-06-04 | 2023-08-01 | Sprinklr, Inc. | Content compliance system |
US11144730B2 (en) | 2019-08-08 | 2021-10-12 | Sprinklr, Inc. | Modeling end to end dialogues using intent oriented decoding |
Also Published As
Publication number | Publication date |
---|---|
US20170068648A1 (en) | 2017-03-09 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US10140646B2 (en) | System and method for analyzing features in product reviews and displaying the results | |
US8700480B1 (en) | Extracting quotes from customer reviews regarding collections of items | |
US20120254060A1 (en) | System, Method, And Computer Readable Medium for Ranking Products And Services Based On User Reviews | |
US20170068965A1 (en) | System and method for analyzing reviews to determine feature score | |
JP6431056B2 (en) | Ranking comments by search engines | |
US9372592B1 (en) | Presenting information related to content items | |
US20140025532A1 (en) | Correlated Content Recommendation Techniques | |
US20140172744A1 (en) | System and method for providing access to product reviews and/or ratings | |
US20100153372A1 (en) | 3d visualization system for web survey | |
Wang et al. | Database submission—market dynamics and user-generated content about tablet computers | |
US8515953B2 (en) | Temporal visualization of query results | |
CN107111640B (en) | Method and user interface for presenting auxiliary content with image search results | |
US10558657B1 (en) | Document content analysis based on topic modeling | |
US20150221023A1 (en) | Information providing device, information providing method, information providing program, and computer-readable storage medium storing the program | |
US10255283B1 (en) | Document content analysis based on topic modeling | |
US20130212039A1 (en) | Review timeline for ownership lifecycle experience | |
EP2778979A1 (en) | Search result ranking by brand | |
US11164223B2 (en) | System and method for annotating reviews | |
CN111782951A (en) | Method and device for determining display page, computer system and medium | |
Chen et al. | Experiment on sentiment embedded comparison interface | |
US20170068975A1 (en) | System and method for displaying reviews according to features | |
Wong et al. | Learning to extract and summarize hot item features from multiple auction web sites | |
Zheng et al. | Substitutes or complements: another step forward in recommendations | |
US20220083617A1 (en) | Systems and methods for enhanced online research | |
US9965470B1 (en) | Extracting quotes from customer reviews of collections of items |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AS | Assignment |
Owner name: WAL-MART STORES, INC., ARKANSAS Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:PARVEEN, TABASSUM;JAIN, SIDDHANTH;MOHAPATRA, HRUSHIKESH;SIGNING DATES FROM 20150914 TO 20150915;REEL/FRAME:036757/0426 |
|
AS | Assignment |
Owner name: WM GLOBAL TECHNOLOGY SERVICES INDIA PRIVATE LIMITE Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:PARVEEN, TABASSUM;JAIN, SIDDHANTH;MOHAPATRA, HRUSHIKESH;SIGNING DATES FROM 20180702 TO 20180705;REEL/FRAME:046280/0291 |
|
STCF | Information on status: patent grant |
Free format text: PATENTED CASE |
|
AS | Assignment |
Owner name: WAL-MART STORES, INC., ARKANSAS Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:WM GLOBAL TECHNOLOGY SERVICES INDIA PRIVATE LIMITED;REEL/FRAME:054787/0768 Effective date: 20191017 |
|
AS | Assignment |
Owner name: WALMART APOLLO, LLC, ARKANSAS Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:WAL-MART STORES, INC.;REEL/FRAME:059036/0629 Effective date: 20220217 |
|
MAFP | Maintenance fee payment |
Free format text: PAYMENT OF MAINTENANCE FEE, 4TH YEAR, LARGE ENTITY (ORIGINAL EVENT CODE: M1551); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY Year of fee payment: 4 |